With advances in technology, genomics is becoming increasingly important for precision medicine, both for determining risk to develop disease and for treatment decisions. The high-risk breast cancer genes, BRCA1 and BRCA2, were discovered in the mid-1990s. Since then, many other risk loci have been identified by genome-wide association and, more recently, by sequencing studies. However, despite these efforts, the combined effects of known breast cancer loci only explain approximately one-third of the genetic risk. Most of the genes predisposing to breast cancer still have not been identified, thus limiting the potential for identification of many individuals at high risk of developing the disease. Whole-exome sequencing is a potentially powerful approach to discover rare mutations in genes that may be associated with breast cancer susceptibility. However, the number of variants discovered in these studies is large and prioritizing them for replication is difficult. Somatic mutations and copy number and other changes in genes in tumors may provide biological insights that can guide gene selection for replication. In particular, many cancer susceptibility genes act by a pathogenic mutation of the germline copy and then somatic inactivation of the second copy. Use of datasets such as The Cancer Genome Atlas (TCGA) may help to identify additional cancer genes by determining which genes display germline and somatic events in the same individual. However, Latinos, African Americans and Asians are under-represented in TCGA, and because rare variants are usually population specific, it is important that populations of all ancestral backgrounds be included in gene discovery. Latinos represent the largest and fastest-growing minority population in the U.S., yet they have been largely under-studied in terms of genetic susceptibility to breast cancer. Although Latinas have an overall lower incidence of breast cancer than non-Latina Caucasians, Latinas are more likely to be diagnosed at a younger age and to have more aggressive disease at diagnosis. Our proposal will use a combined germline and somatic genetics approach to discover breast cancer susceptibility genes in Latinas. In Specific Aim 1, we will identify mutations in genes from whole-exome sequencing of 1,400 Latina breast cancer cases diagnosed under age 50 and/or with a family history and 1,400 healthy Latina controls. We will perform gene-based association tests to prioritize genes based on p-value for more testing in Aim 2 and for replication in Aim 3. In Specific Aim 2, we will perform targeted sequencing and expression analysis of 800 genes from Aim 1 in DNA and RNA from 384 breast tumors of cases from Aim 1. In Specific Aim 3, we will use custom capture to target and sequence the top 1,000 candidate genes in 2,300 Latina breast cancer cases and 2,300 controls. At the end of this project, we will have identified a set of genes that are significantly associated with breast cancer in Latinas. A long-term goal is to develop a clinical genetic test that can help to determine risk in individual Latinas who have not been diagnosed to aid in their decision-making with regard to preventive options.